Neuro Oligo Platform (NOP)

A SimBiology quantitative systems pharmacology (QSP) model of oligonucleotide PK/PD across plasma, brain (superficial/deep), CSF, spinal cord and liver compartments. This repository contains the model, data, and driver scripts used to reproduce the manuscript figures.

Requirements

• MATLAB (R2024b was used)
• SimBiology — model construction and simulation
• Statistics and Machine Learning Toolbox — summary statistics used by the plotting helpers (and partialcorr/makedist for the sensitivity analysis)
• Parallel Computing Toolbox — optional. The scripts set args.UseParallel = 1; if you do not have this toolbox, set args.UseParallel = 0 before running (or edit it in the script).

All required input files (args_NOP.xlsx, parameters/, and the curated datasets in data/) are included in the repository.

Setup

1. Clone/download the repository and open MATLAB.
2. In MATLAB, set the Current Folder to this project folder (IT_NABM).
3. When you press Run on a script for the first time, if MATLAB prompts, choose "Change Folder" so the current folder is this project folder. The scripts add the repository to the MATLAB path automatically (addpath(genpath(pwd))).

Reproducing the manuscript figures

Open each script below in the MATLAB Editor and press Run (▶) or F5. Each script builds the model, loads the data, and generates its figures. Figures are displayed on screen and exported to the figures/ folder (configurable in args_NOP.xlsx).

1. PK/PD figures — simulate_NOP.m

Builds the model and runs the BIIB080, APP siRNA (Mivelsiran), Tominersen, and Jacifusen scenarios (two translation variants each), producing the model-vs-data PK/PD overlay figures.

2. Sensitivity analysis (one-at-a-time) — SA_main.m

Runs one-at-a-time parameter sweeps over the brain oligo half-life, brain protein half-life, IC50, cellular uptake half-life, and mRNA half-life, producing the local sensitivity figures.

3. Allometric scaling — scaling.m

Standalone script (base MATLAB only). Produces the cross-species allometric scaling figures and saves them as IC50_final.png, oligo_hl_final.png, and prot_hl.png in the project folder.

4. Global sensitivity analysis (PRCC) — calibration/script_NOP_calibration.m

Runs the global sensitivity analysis and produces the PRCC global sensitivity figure (and Sobol figures). By default the parameter-estimation step is off (args.switches.calibration = 0) and PRCC/Sobol are on, so simply pressing Run generates the sensitivity figures. This script locates the project root automatically and can be run regardless of the current folder.

Notes

• All exported figures go to the figures/ folder by default; the location and file format are set in args_NOP.xlsx.
• The first run is slower because the model is compiled and (if enabled) a parallel pool is started.
• calibration/script_NOP_calibration.m (re)writes a large results file, calibration/PSA_results_NOP.mat, which is not tracked in version control.

License

Please see Genentech Non-Commercial Software License.docx for license